Comments on Stefan Gerlach and Thomas J. Jordan, Tactics and Strategy in Monetary Policy: Benjamin Friedman s Thinking and the Swiss National Bank * Lars E.O. Svensson Sveriges Riksbank, Stockholm University, CEPR, and NBER First draft: April 2011 This draft: October 2011 I am very happy to have the opportunity to participate in this conference and to honor Ben Friedman s work. Ben deserves to be honored. He is not only a great economist; he is also a wonderful person. The Gerlach and Jordan paper is organized around a great idea, namely comparing Ben s thinking about the tactics and strategy of monetary policy with the practical monetary policy of the Swiss National Bank. In line with this, the paper first discusses money and monetary targeting and Ben s thinking about monetary targeting. As you know, Ben is a long-time skeptic of monetary targeting and has provided some of the most insightful criticism of this monetary-policy regime. The paper goes through the SNB s somewhat problematic experience of monetary targeting 1975-1999 and its eventual abandonment of this regime in 1999. Next, the paper discusses Ben s thinking about price stability and inflation targeting and compares this with the SNB s current framework, where the prominent elements are a definition of price stability, a broad-based inflation forecast, and a target for the three-month Libor interest rate. The paper also discusses monetary policy and the recent crisis, the SNB s experience, and the lessons from the crisis that Ben has summarized. Although the SNB s historical experience is very interesting and we can learn many useful lessons from it, I will use my limited time to focus on Ben s thinking about inflation targeting. As noted by Gerlach and Jordan, Ben has been quite critical of inflation targeting. In understanding that criticism, I think it is important to make a distinction between strict and flexible inflation targeting. With strict inflation targeting, the only objective is to stabilize inflation around the inflation target. Inflation is the only target variable. There is no regard to the real economy except to the extent that it affects inflation. With flexible inflation targeting, the objective is to stabilize both inflation around the inflation target and resource utilization around a sustainable level. Thus, there are two target variables, inflation and resource utilization. My understanding is that Ben s criticism is directed towards the idea that inflation should be the only target variable, that is, towards strict inflation targeting. Furthermore, to the extent that inflation targeting also has implicit real objectives, it risks becoming opaque and allows the central bank to avoid accountability (Friedman 2008a, p. 59-60). However, this is a situation in which there is not explicit but * This comment was prepared for Recent Developments in Monetary Policy, Fiscal Policy, and Financial System Design: A Conference Honoring Benjamin M. Friedman, held at the Federal Reserve Bank of Boston, April 22-23, 2011. The views expressed here are my own and are not necessarily shared by the other members of the Riksbank s Executive Board or the Riksbank s staff. I thank Riksbank staff member Björn Lagerwall for research assistance. 1
implicit inflation targeting; it assumes that the flexible inflation targeters are not explicit about the realeconomy objectives. But flexible inflation targeters are becoming increasingly transparent. Although several inflation-targeting central banks have become more explicit about their real objectives, Norges Bank and the Riksbank are arguably ahead. Each Monetary Policy Report from Norges Bank starts with the following two points: Objective: The operational target of monetary policy is low and stable inflation, with annual consumer price inflation of approximately 2.5% over time. Implementation: Norges Bank operates a flexible inflation targeting regime, so that weight is given to both variability in inflation and variability in output and employment. Each Monetary Policy Report from the Riksbank starts by stating that the inflation target is 2 per cent and that the Riksbank, in addition to stabilizsing inflation around the inflation target, [is] also striving to stabilise production and employment around long-term sustainable paths. The Riksbank therefore conducts what is generally referred to as flexible inflation targeting. Furthermore, it states that [a] wellbalanced monetary policy is normally a question of finding an appropriate balance between stabilising inflation around the inflation target and stabilising the real economy. I believe this shows that Norges Bank and the Riksbank are pretty transparent about their real objectives. The Riksbank has a six-member Executive Board that makes decisions about the Bank and monetary policy. The members have individual accountability. There are normally six monetary-policy meetings per year. The day after the policy meeting, a press release is published that summarizes the decision and reports the votes and any dissents, a Monetary Policy Report or a Monetary Policy Update is published, and a press conference is held. The Report and Update include a published forecast for inflation, the real economy, resource utilization, the exchange rate, and the policy rate. In the Report, there are also forecasts of inflation and resource utilization for alternative policy-rate paths. Two weeks after the policy meeting, attributed minutes are published. Everything is published in both Swedish and English. There is considerable discussion of resource utilization in the minutes, although the emphasis may vary a bit among the members. Figure 1 shows forecasts for the repo rate (the policy rate), CPIF inflation (CPI inflation excluding the direct effects of policy-rate changes) and unemployment that were published in the February 2011 Monetary Policy Report. Forecasts for three alternative policy-rate paths are shown. It also shows corresponding mean squared gaps for the three policy alternatives. The mean squared gaps provide a measure of the stability of inflation and unemployment. The mean squared gap for inflation is the sum of the squared inflation gaps (the deviations of the inflation forecast from the inflation target) over the forecast period, divided by the forecast horizon (normally 12 quarters). The mean squared gap for unemployment is the sum of the squared unemployment gaps (the deviations of the unemployment rate from the sustainable unemployment rate in this figure, set at 6.5 per cent) over the forecast period, divided by the forecast horizon. 1 According to the mean squared gaps, the lower policy-rate path stabilizes both inflation and the unemployment gap better than the other policy-rate paths. Nevertheless, a majority of four Board members voted in favor of the main-scenario repo-rate path, with arguments reproduced in the minutes (Sveriges Riksbank 2011). 1 For details on mean squared gaps, see Svensson (2011a). 2
Two Board members, Karolina Ekholm and me, dissented in favor of a lower policy-rate path. A lower policy-rate path would be preferable already under the assumptions underlying figure 1. However, we also consider a lower foreign policy-rate forecast, in line with foreign implied market forward rates, to be more appropriate than the higher foreign policy-rate forecast adopted by our colleagues and underlying figure 1. Furthermore, I consider 5.5 per cent, rather than 6.5 per cent, to be the appropriate estimate of the sustainable unemployment rate. The reasons for this are detailed in the minutes. Figure 2 shows forecasts under these assumptions for inflation and unemployment for two policy-rate paths, the main-scenario path of our colleagues and a lower policy-rate path that we dissented in favor of. The minutes provide further details on the debate about the policy decision. Figure 1. Monetary policy alternatives, February 2011 (from the Monetary Policy Report February 2011, except panel b). Sustainable unemployment rate 6.5 per cent. 3
Figure 2. Monetary policy alternatives, February 2011 (from the minutes of the February 2011 meeting). Foreign interest rates according to market expectations; sustainable unemployment rate 5.5 per cent. Ben has repeatedly defended the Federal Reserve s dual mandate of price stability and maximum sustainable employment. Let me finally make the point that I don t see any fundamental differences between flexible inflation targeting and the dual mandate. For instance, both are consistent with a standard quadratic loss function. The publication of Federal Open Market Committee longer-run projections for inflation and the unemployment rate makes it possible to identify the mandate-consistent inflation rate, the inflation rate that the FOMC considers to be consistent with its mandate and the analogue of an inflation target, and the FOMC s estimate of the longer-run sustainable unemployment rate. 2 Indeed, I believe there is sometimes a misunderstanding about hierarchical and dual mandates. In contrast to Meyer (2004), for instance, I maintain that there is no fundamental difference between the dual mandate of the Federal Reserve and the hierarchical mandate of the Riksbank (which is consistent with the statutes of the European System of Central Banks). 3 The Riksbank s mandate is considered hierarchical because price stability is its primary objective. But here we have to distinguish between first and second moments, that is, means and variances. The hierarchical mandate applies to the first moments, the unconditional means. Regarding the unconditional mean and the average level of inflation, there is an explicit inflation target (an inflation rate of 2 per cent per year) that is chosen and enforced by the Riksbank. There is no explicit target level for resource utilization that is chosen by the Riksbank. Instead, what is a sustainable level of resource utilization is determined by the properties of the economy and structural policies and is estimated by the Riksbank. Thus, the Riksbank has an independently chosen target for inflation but no independently chosen target for output, employment, unemployment, or any other resource utilization- 2 The similarity of the Federal Reserve s dual mandate and the Riksbank s flexible inflation targeting is further discussed in Svensson (2011b). This paper also compares Federal Reserve and Riksbank policy in June/July 2010. 3 See the discussion of Meyer (2004) in Svensson (2004). 4
related variable. For the second moments, the variability of inflation and resource utilization, the Riksbank has a dual mandate in that it aims to stabilize both inflation around the chosen target and resource utilization around the estimated normal level. Looked at this way, there is no fundamental difference between a hierarchical mandate and a dual mandate. Indeed, in a discussion of dual and hierarchical mandates, Ben notes that (Friedman 2008b, p. 158) [a] hierarchical instruction to the central bank is, really, a kind of dual mandate as well. The difference is that it is an unspoken one. Thus, the problem is whether the real-economy objectives are transparent or not. Ben is clearly not a skeptic of flexible inflation targeting, as long as the real-economy objectives are sufficiently transparent. References Friedman, Benjamin M. (2008a), What Objectives for Monetary Policy, in Touffut (ed.), Central Banks as Economic Institutions, Edward Elgar, 52-67. Friedman, Benjamin M. (2008b), Why a Dual Mandate is Right for Monetary Policy, International Finance 11:2, 153-165. Gerlach, Stefan, and Thomas J. Jordan (2011), Tactics and Strategy in Monetary Policy: Benjamin Friedman s Thinking and the Swiss National Bank, working paper. Meyer, Laurence H. (2004), Practical Problems and Obstacles to Inflation Targeting, Federal Reserve Bank of St. Louis Review 84(4) (July/August 2004) 151-160. Svensson, Lars E.O. (2004), Commentary (on Laurence H. Meyer, Practical Problems and Obstacles to Inflation Targeting"), Federal Reserve Bank of St. Louis Review 84(4) (July/August 2004) 161-164. Svensson, Lars E.O. (2011a), Inflation Targeting, in Friedman, Benjamin M., and Michael Woodford, eds., Handbook of Monetary Economics, Volume 3b, chapter 22, Elsevier. Svensson, Lars E.O. (2011b), Practical Monetary Policy: Examples from Sweden and the United States, Brookings Papers on Economic Activity, Fall 2011, forthcoming. Sveriges Riksbank (2011), Minutes of the Executive Board s Monetary Policy Meeting, No.1, February 14, 2011, www.riksbank.se. 5